云原生家庭网络（十七）：使用 nftables 透明拦截流量

        lifecycle:
          postStart:
            exec:
              command:
              - /scripts/set-rules.sh
          preStop:
            exec:
              command:
              - /scripts/clean.sh

        volumeMounts:
        - mountPath: /scripts
          name: script
        - mountPath: /etc/nftables
          name: nftables-config

      volumes:
      - configMap:
          defaultMode: 420
          name: nftables-config
        name: nftables-config
      - configMap:
          defaultMode: 511
          name: script
        name: script

      initContainers:
      - command:
        - /scripts/clean.sh
        image: your-proxy-image
        imagePullPolicy: IfNotPresent
        name: clean
        securityContext:
          privileged: true
        volumeMounts:
        - mountPath: /scripts
          name: script

#!/usr/sbin/nft -f

table inet proxy
delete table inet proxy

table inet proxy {

    # 保留网段，参考：https://zh.wikipedia.org/zh-sg/%E4%BF%9D%E7%95%99IP%E5%9C%B0%E5%9D%80
    set byp4 {
        typeof ip daddr
        flags interval
        elements = {
            0.0.0.0/8,
            10.0.0.0/8,
            100.64.0.0/10,
            127.0.0.0/8,
            169.254.0.0/16,
            172.16.0.0/12,
            192.0.0.0/24,
            192.0.2.0/24,
            192.88.99.0/24,
            192.168.0.0/16,
            198.18.0.0/15,
            198.51.100.0/24,
            203.0.113.0/24,
            224.0.0.0/4,
            240.0.0.0-255.255.255.255
        }
    }
    set byp6 {
        typeof ip6 daddr
        flags interval
        elements = {
            ::,
            ::1,
            ::ffff:0:0:0/96,
            100::/64,
            64:ff9b::/96,
            2001::/32,
            2001:10::/28,
            2001:20::/28,
            2001:db8::/32,
            2002::/16,
            fc00::/7,
            fe80::/10,
            ff00::/8
        }
    }

    chain prerouting {
        type filter hook prerouting priority filter; policy accept;

        meta l4proto != { tcp, udp } counter return

        # 避免拦截公网回包（如果没用主路由方案，可以删除）
        iifname ppp0 \
          ct state new \
          ct mark set 11 \
          log prefix "Public Inbound:" \
          counter \
          return \
          comment "Mark conntrack for connection from internet, avoid intercept the return packets"

        # 不拦截从公网主动进来的包（如果没用主路由方案，可以删除）
        iifname ppp0 counter return

        # 避免源和目的都是本机公网 IP 时造成代理 UDP 无限回环
        fib daddr type local counter return

        # 对保留 IP(包括内网) 放行。
        ip daddr @byp4 counter return
        ip6 daddr @byp6 counter return
        goto proxy
    }
    chain proxy {
        # 对其它数据包(其它内网机器访问公网的包)进行拦截，全部转发给代理。
        # 踩坑：一定 tproxy 到回环 ip (127.0.0.1和::1)，如果 tproxy 到 0.0.0.0和::，运行一段时间后，可能会导致 UDP 疯狂不停回环打满 CPU（本机外部IP到本机外部IP一直在lo网卡打转）
        meta l4proto { tcp, udp } tproxy ip to 127.0.0.1:12345 meta mark set 1 counter
        meta l4proto { tcp, udp } tproxy ip6 to [::1]:12345 meta mark set 1 counter
    }
    chain output {
        type route hook output priority filter; policy accept;

        # 不拦截代理自身发出的包。
        meta skgid eq 23333 counter return

        # 不拦截公网回包
        ct mark 11 counter return

        # 对保留 IP(包括内网) 放行。
        ip daddr @byp4 counter return
        ip6 daddr @byp6 counter return

        # 避免源和目的都是本机公网 IP 时造成代理 UDP 无限回环
        fib daddr type local counter return

        # 主机内进程发出的包打上 mark，让报文走策略路由重新进入 PREROUTING，以便被代理。
        meta l4proto { tcp, udp } meta mark set 1 counter
    }
}

#!/bin/bash

set -x

while :; do
	export https_proxy=http://127.0.0.1:10809                                           # 假设代理会启动 10809 端口的 HTTP 代理
	code=$(curl -I -o /dev/null -m 5 -s -w %{http_code} https://www.google.com | xargs) # 通过代理探测目标地址是否正常响应
	if [[ "${code}" == "200" ]]; then                                                   # 探测成功就退出探测循环，准备设置策略路由
		break
	fi
	echo "bad code:${code};" >/tmp/error.log
	sleep 5s
done

echo "proxy is ready, set up rules"
ip rule list | grep "from all fwmark 0x1 lookup 100"
if [ $? -ne 0 ]; then
	echo "add rule and route"
	ip rule add fwmark 0x1 table 100
	ip route add local 0.0.0.0/0 dev lo table 100
	ip -6 rule add fwmark 0x1 table 100
	ip -6 route add local ::/0 dev lo table 100
fi

nft -f /etc/nftables/nftables.conf # 设置 nftables 规则
exit 0

#!/bin/bash

set -x

nft delete table inet proxy # 关闭时清理 nftables，避免拉不了镜像
ip rule delete fwmark 0x1 table 100
ip route delete local 0.0.0.0/0 dev lo table 100
ip -6 rule delete fwmark 0x1 table 100
ip -6 route delete local ::/0 dev lo table 100

exit 0

log-queries=extra
no-resolv
no-poll
server=61.139.2.69
strict-order
log-dhcp
cache-size=2000
dhcp-range=10.10.10.11,10.10.10.254,255.255.255.0,12h
dhcp-authoritative
dhcp-leasefile=/var/lib/dnsmasq/dnsmasq.leases
dhcp-option=option:router,10.10.10.2
dhcp-option=option:dns-server,61.139.2.69,218.6.200.139
dhcp-host=a0:78:17:87:0e:51,10.10.10.5 # M1
dhcp-host=fc:18:3c:37:6a:1c,10.10.10.6 # iPhone - roc